Thanks for the quick reply. I meant mdb_txn_commit (instead of mdb_txn_close), sorry about that.
> If there was a restriction on usage in read-only txns the docs would have > said so. This is what I guessed too. However I now have a follow-up question: The documentation for mdb_txn_abort() says: "The transaction handle is freed. It and its cursors must not be used again after this call, except with #mdb_cursor_renew()." Could it be worth it to change this to "...except with #mdb_cursor_renew() or #mdb_cursor_close()"? This would have cleared the ambiguity for me (and I would never have created this thread to begin with). Regards, Sam Jan 24, 2022, 17:43 by [email protected]: > Sam Dave wrote: > >> Hello, >> >> >> The mdb_cursor_close() documentation says: "Its transaction must still be >> live if it is a write-transaction." >> >> >> This leads me to my question: For *read-only* transactions, is it allowed to >> run mdb_cursor_close() *after* mdb_txn_close() or mdb_txn_abort()? >> > > There is no mdb_txn_close(). > > If there was a restriction on usage in read-only txns the docs would have > said so. > > -- > -- Howard Chu > CTO, Symas Corp. http://www.symas.com > Director, Highland Sun http://highlandsun.com/hyc/ > Chief Architect, OpenLDAP http://www.openldap.org/project/ >
